SUMMARY/ABSTRACT In countries with high tuberculosis (TB) incidence, TB is the most common cause of pericardial effusion. Pericardial tuberculosis (PCTB) is a severe form of extrapulmonary TB causing hemodynamic instability and cardiac tamponade requiring pericardiocentesis3. Despite anti-tuberculosis treatment (ATT), complications remain frequent and mortality high (8-34%, worst in HIV co-infected persons). Death in PCTB is contributed to both by a high pericardial fluid (PCF) bacillary load that may arise from inadequate penetration of ATT into pericardium, and by dysregulation of the immune response, which attempts to clear the infection. Thus there is an important need to improve the current PCTB treatment approach both in terms of efficiency of bacterial killing and identification of deleterious inflammatory pathways that could be specifically modulated. This collaboration will therefore determine: 1: Whether high dose rifampin increases drug exposure in PCF and thus results in a greater decline in Mtb load. The hypothesis is that rifampin (RIF) exposure in PCF will increase with a higher dose (35mg/kg, RIF35) versus standard (RIF 10mg/kg, RIF10) and that decline in bacteria will correlate with RIF exposure in PCF and plasma. A randomized controlled clinical trial of 35mg/kg RIF daily in combination with standard doses of the 3 other antitubercular drugs (pyrazinamide, ethambutol and isoniazid), versus standard of care will be conducted. This will assess safety and the primary efficacy endpoint of PCF bacillary burden at 0 and 72 hrs. Secondary endpoints will be a composite of clinical measures and precise cardiovascular magnetic resonance (CMR) evidence of pericardial inflammation, thickening, effusion, fibrosis or constriction. The pharmacokinetics and pharmacodynamics of antitubercular drugs in PCF will also be determined. 2: Relationships between pericardial Mtb-specific T cells with bacterial load and treatment outcome in PCTB. Multiple types of T cells are found in PCF, but it remains unknown which subsets contribute to protection and/or pathology. The phenotype and function of conventional and donor unrestricted (e.g. MAIT cells, CD1- restricted T cells, ?? T cells) Mtb-specific T cell subsets in paired blood and PCF from HIV infected and uninfected persons with PCTB will be investigated. The relationship in PCF of CD4+ T cell biomarkers (CD153, HLA-DR, CD38, Ki67) with bacillary load and outcome will also be undertaken. 3: Relationships between Mtb-induced markers of host cell death pathways and bacterial load in PCTB. In model systems, necrotic (cytolytic) cell death benefits dissemination of Mtb, while apoptotic cell death associates with host bacterial control. It is unknown how the induction of varying cell death pathways relate to disease in TB patients. Therefore well-defined markers of cytolytic macrophage cell death pathways in PCF (and blood) will be determined and their correlation with clinical endpoints explored.